ANALYTICA

CHIMICA
ACTA
ELSEVIER Analytica Chimica Acta 333 (I 996) I I S-l 18

Determination of tin in canned fruit juices by stripping

potentiometry
R. Ratana-ohpas*, P. Kanatharana, W. Ratana-ohpas, W. Kongsawasdi

Received 7 March 1996; accepted 29 April 1996

Abstract

The total tin concentration in five different kinds of canned fruit juices has been determined by using constant current
stripping analysis on a gold film electrode, and potentiometric stripping analysis on a mercury film electrode, the
concentration in the samples ranging between 40 and XOmg I--‘. The total time for one determination, including sample
preparation, is approximately 10 min and the relative precision around 8%. The results obtained by the two electroanalytical
techniques agreed satisfactorily with those results obtained by flame atomic absorption spectrophotometry.

Kqwwd.v: Tin; Canned fruit juices; Potentiometry; Constant current stripping analysis: Stripping potentiometry; Heavy metal: Beverage;
Trace element analysis

1. Introduction electrodes, and since elemental tin is very soluble in

these elements, electroanalytical stripping methods
Canned food frequently contains high concentra- might be alternatives to flame atomic absorption
tions of tin. This is particularly true for acid food- spectrophotometry as proposed earlier by Mannino
stuffs such as canned fruit juices. Although tin is not [4,5]. In this article two different electroanalytical
a highly toxic element, high concentrations of tin, in approaches to the determination of total tin in canned
the order of 0.1 and 1.Og lP’, affects the flavour of the fruit juices, constant current stripping analysis [6] on
juice and can also cause diarrhoea. For this reason a gold film electrode and potentiometric stripping
several methods have been proposed for the determi- analysis [7] on a mercury film electrode, are des-
nation of tin in canned fruit juices, including cribed. The results are compared with those obtained
gravimetric, titrimetric and calorimetric methods by flame atomic absorption spectrophotometry.
[ I-3). Flame atomic absorption spectrophotometry
is, no doubt, by far the most commonly used method. 2. Experimental
Since both tin(H) and tin(IV) can be electroche-
mically reduced to elemental tin on mercury and gold 2.1. Instrumentation and electrodes

Constant current stripping analysis and potentio-

* Corresponding author. metric stripping analysis were performed with a

0003-2670/96/$15.00 ~‘8 1996 Elsevier Science B.V. All rights reserved

PII SOOO1-2670(96)002 16-4
116 R. Ratana-ohpas et al./Analytica Chimica Acta 333 (1996) I IS-118

TraceLab 20 System (Radiometer Analytical SA, was checked and, if the calibration deviated by more
Lyon, France) in combination with an IBM compa- than 5% from the previous calibration, the instrument
tible personal computer. Fully automated analysis was recalibrated.
was achieved using the TAP2 TraceTalk: Method
Builder and Commander software package provided 2.4. Potentiometric stripping analysis (PSA)
by the instrument manufacturer. A 3 mm diameter using mercury film electrode
glassy carbon disc electrode was used as the substrate
for the mercury or gold films in combination with a A mercury film was plated onto the glassy carbon
saturated calomel reference electrode (SCE) and a electrode by stirred electrolysis for 15 s at -0.70 V
platinum counter electrode. The flame atomic absorp- vs. SCE in a solution containing 500mgl-’ mercur-
tion measurements were performed with a Shimadzu ~(11) in 1 M hydrochloric acid. Subsequent to rinsing
AA 680 instrument. in distilled water tin was determined in the samples,
pre-treated as mentioned above, by stirred electro-
2.2. Samples and sample preparation lysis at -1.OOV vs. SCE for 60 s prior to stripping
without any applied constant current, i.e. by using
Five different kinds of canned fruit juices, guava, both 20 mg 1-l of Hg(I1) and dissolved oxygen in the
lychee, tomato, mango and pineapple were obtained sample as the oxidants. Calibration was performed in
from local department stores. Prior to analysis 2 ml of the same way as for constant current stripping
fruit juice samples were mixed with 20 ml 1 : 1 (v/v) analysis.
concentrated nitric acid/hydrochloric acid. To 0.50 ml
of this mixture 10 ml 6 M hydrochloric acid was then
added and the total volume finally adjusted to 30ml 3. Results and discussion
using doubly distilled water. This solution was used
for the constant current stripping determinations 3.1. Constant current stripping analysis
using gold film electrodes and for the atomic
absorption measurements. In case of potentiometric Fig. 1 shows the stripping curves obtained in the
stripping analysis 0.60 ml of 1000 mg l- ’ of mercur- analysis of standard solutions containing 50, 100 and
~(11) was also added to the solution before the total 150 ugll’ (ppb) of tin(IV) using constant current
volume finally adjusted to 30ml using doubly stripping on a gold film electrode. Also shown is the
distilled water. stripping curve obtained in the analysis of a diluted
sample solution of lychee juice. The stripping signals,
2.3. Constant current stripping analysis expressed in milliseconds, are also shown in the
(CCSA) using gold film electrode figure. As can be seen from Fig. 1 tin gives two
stripping peaks on the gold film electrode, one at a
A gold film was plated onto the glassy carbon potential of approximately -0.2OV vs. SCE and
substrate by stirred electrolysis for 30s in a solution another at approximately +0.20 V. The mechanisms
containing 500 mg 1-l gold(II1) in 1 M hydrochloric involved in the two peaks have not been fully
acid at a potential of -0.10 V vs. SCE. The electrode elucidated, the area of both being proportional to
was then rinsed with distilled water and the juice the sample concentration of tin(IV). The most
samples, pre-treated as described above, were then obvious explanation, as has also been proposed
analysed for tin by stirred electrolysis for 30s at by several investigators, is that the first peak
-0.7OV vs. SCE prior to stripping with a constant corresponds to the oxidation of elemental tin to
current equal to 15 PA. The tin concentration was tin(I1) and the second peak the oxidation of tin(I1) to
finally evaluated from a single point calibration in a tin(IV). Although separate experiments, using
solution containing 200 pgl-’ tin(IV) and the same coulometric stripping potentiometry [8], has revealed
concentration of mineral acids as the samples. The that the second peak is a two-electron reaction the
same calibration was used for several determinations. first peak seems to correspond to approximately one
Approximately for every fifth analysis the calibration electron [9]. In addition the compound formed in
 R. Ratana-ohpas et ul./Analytrcn Chimica Ac,ttr 333 (19% I 115-l 18 117

Table I
Results obtained from five consecutive determinations of tin in
three different standard solutions of tin(IV) by constant current
<tripping analysis (CCSA) and by potentiometric stripping unaly\is
(PSA)

Sn(IV) taken l~gl ’ SnUV) found I fl (rugI ’I

CCSA PSA

50.0 5351 I .2 51.7_to.7

IlJV llPJ
100.0 99.6Lt I.3 Ul.LH2.7
.400, 100 ppb S” .400 150 ppb S”
i 150.0 143.710.X I s-I.41I .9
*200 171.60c! *zoo 260.233

01+== 0I=====-
I
-200 i;' 73067

-I -200L
0 10 siV 0 10 S/V

Fig. I. Stripping potentiograms from the constant current strip-

ping analysi\ (CCSA) measurements of tin in three different
standard holutions of tin(N) 50, 100 and I50 ygl-’ (ppb), and a
I :660 diluted sample solution of lychee juice using gold tilm
electrode.

the first oxidation step must be a solid remaining

on the electrode surface and thus available quantita-
tively for a second oxidation step. If soluble
Fig. 2. Stripping potentiograms from the potentiometric ctrippmg
species had been formed, these would have
analysis (PSA) measurements of tin in three different standard
diffused away from the electrode surface and solutions of tin(W): 50. 100 and ISOpgI ‘, and a I :660 diluted
only a minor, and irreproducible, fraction remaining sample solution of lychee juice using mercury lilm electrode.
for the second oxidation step. In this investigation the
second stripping peak was used for the quantification Table 2
Comparison between the results obtained from live consecutive
of tin in diluted solutions of canned fruit juice
determinations of tin in five different canned fruit juices by
samples. Table 1 shows the results obtained in the
constant current strIpping analysis (CCSA). potentiometric strip-
analysis of three different standard solutions of ping analysis (PSA), and flame atomic absorption spectrophoto-
tin(W). each standard solution being analysed five metry (FAAS)
times. Samples Sn(lV) foundfn (mg I ’I

3.2. Pr~tentiometric stripping analy.sis CCSA PSA FAAS

Guava ,juicc 54.2* I .2 39.x+ I .h 59.4*2.3

Fig. 2 shows the potentiometric stripping curves Lychee juice 43.931.8 45.3r7.i 45.7 t.3.x

obtained in the analysis of standard solutions Tomato juice 58.7% IA 6Y.7iO.5 65.21-1.3
Mango juice h3.3k2. I 77.51 I .h 73.5 t2.9
containing 50, 100 and 150 pgl-’ of tin(IV) and
Pineapple juice 57.x+0.9 65.5i1.7 6Y. IS.8
in the analysis of a diluted sample solution of
lychee juice (cf. Fig. 1). On the mercury film
electrode tin gives only one stripping peak, the peak In Table 2 the results obtained from potentiometric
potential being approximately -0.55 V vs. SCE. stripping analysis of five different canned fruit juices
Table I shows the results from five consecutive are compared with the results obtained by flame
determinations of the three different standard solu- atomic absorption spectrophotometry and constant
tions of tin(N). current stripping analysis.
118 R. Ratana-ohpas et al./Analytica Chimica Acta 333 (1996) 115-118

3.3. Precision, detection limit and accuracy monitoring of tin in canned fruit juices. Specific
advantages are the relative simple sample prepara-
As can be seen from Table 1 there is no significant tion, low instrumental cost and compact size and no
difference in precision between potentiometric strip- need for compressed gases or ventilation system. In
ping analysis and constant current stripping analysis certain applications the portability of the instrument
in the analysis of tin(N) standards. The relative might be an additional advantage.
precision for both techniques can be estimated as 4-
6%. From the stripping time values shown in Fig. I
and 2 it is seen that the sensitivity is slightly lower for Acknowledgements
potentiometric stripping analysis than it is for
constant current stripping analysis, 1.0 and This work was supported by the International
1.7mslpg~‘, respectively. Assuming an instrumen- Program in the Chemical Sciences (IPICS), Uppsala,
tal detection limit of 1 ms this corresponds to Sweden and partially by International Development
detection limits of 1.0 and 0.6pgll’ of tin(IV). Program (IDP), Australia. The work was performed
Since the tin(IV) concentrations in the present in close co-operation with the Department of
application are much higher than the detection limits Analytical and Marine Chemistry (AMK), University
the difference in sensitivity between the two of Goteborg and Chalmers University of Technology,
techniques is of no practical significance. Goteborg, Sweden. Special thanks are due to
Table 2 shows that there is a satisfactory correla- Professor Daniel Jagner, Dr. Lars Renman and Dr.
tion between the results from the two different Yudong Wang for introducing the authors to electro-
electroanalytical techniques and the results obtained analytical chemistry and for continuous support.
by flame atomic absorption spectrophotometry. There
is no significant difference between the potentio-
metric stripping analysis results and the spectro- References
photometric results. The constant current stripping
analysis results are, however, significantly lower with [ll Association of Official Agricultural Chemists, Official
Methods of Analysis, 10th edn., Washington, 1965, p, 380.
approximately 8%, the reason for this is not fully
121 D. Pearson, The Chemical Analysis of Food, 5th Edn., J.& A.
understood. Churchill, London, 1962.
[31 K. Liska, Anal. Abstr., 3 (1956) 3590.
[41 S. Mannino, Analyst, 108 (1983) 1257.
4. Conclusions [51 S. Mannino, Analyst, 109 (1984) 905.
[61 H. Huiliang, D. Jagner and L. Renman, Anal. Chim. Acta,
202 (1987) 117.
Constant current stripping analysis and potentio- D. Jagner and A. Graneli, Anal. Chim. Acta, 83 (1976) 19.
[71
metric stripping analysis are attractive alternatives to 181 D. Jagner, Y. Wang and F. Ma, Electroanalysis, 7 (1995) 614.
flame atomic absorption spectrophotometry for the L9l D. Jagner and E Ma, personal communication.